ChatGPT Has Been Leaking Your Conversations

According to OpenAI CEO Sam Altman, a ChatGPT malfunction enabled some users to view the titles of other users’ conversations. He expressed regret over the “significant” mistake but confirmed that it had been resolved.

Users posted screenshots of chat logs on social media sites like Reddit and Twitter, claiming that they were not their own. Despite the fix, many users remain concerned about the platform’s privacy.

Since its debut in November last year, ChatGPT has been utilized by millions of individuals to create messages, compose music, and even program. Every interaction with the chatbot is saved in the user’s chat history, which can be viewed later.

But earlier this week, users started seeing conversation histories in their accounts that they never had with the chatbot.

A Reddit user posted an image of their chat history, which included titles such as “Chinese Socialism Development” and conversations in Mandarin.

On Tuesday, OpenAI informed Bloomberg that they had briefly deactivated the chatbot to correct the problem. The company also stated that users were unable to access the actual conversations at that time.

The recent glitch has raised concerns about OpenAI’s access to people’s private conversations, leading to worries about potential data breaches. OpenAI’s CEO announced on Twitter that a “technical postmortem” would be conducted soon to investigate the matter.

While the company’s privacy policy states that user data, including prompts and responses, may be utilized for model training, such data is only accessed once all personally identifiable information has been removed.